Mammalian semen contains approximately equal numbers of Y-chromosome bearing sperm cells (Y-sperm) and X-chromosome bearing sperm cells (X-sperm). Fertilization of an ovum by a Y-sperm produces a male. Fertilization by an X-sperm produces a female.
Various methods have been proposed for modifying mammalian semen to increase the relative percentage of X- or Y-sperm in a semen sample, and thereby achieve a greater likelihood of female or male offspring. Attempts to influence or control mammalian sex have not been verifiable. (For reviews of prior research, see Garner, 1984; Pinkel et al., 1985.)
One of the more common approaches for attempting X-sperm or Y-sperm enrichment in semen has relied on motility and density sedimentation. (See Kaiser et al., 1974, or Soupart, 1975.) This approach is based on the Y-sperm's purported greater motility and lighter weight than X-sperm. According to the theory, Y-sperm would penetrate an interface created at two different media densities more easily than X-sperm. One such approach used albumin gradient sedimentation. However, due to the morphological variability of the maturing sperm, no one has independently shown that this technique can separate or enrich X-sperm or Y-sperm (Brandriff et al., 1986).
Immunological methods have also been tried as a means of separating X- and Y-sperm. These methods are based on the fact that spermatid RNA polymerase is capable of transcribing the haploid genome (Moore, 1971). It was believed that X- and Y-sperm could be separated immunologically on the basis of the different antigens produced from this RNA transcript. Antigens investigated in unsexed sperm included the LDH isozyme (Stambaugh and Buckley, 1971). Again, no demonstrable separation has been reported.
Investigators have also looked to the male H-Y antigen as a potential means to enrich sperm subpopulations and thereby preselect sex of offspring. Indirect evidence suggested that H-Y antigen was a cell-surface antigen produced in males but not in females. Accordingly, investigators have reasoned that H-Y is expressed by cells containing a Y chromosome and, therefore, on the surface of Y-sperm but not X-sperm. Consequently, some investigators believed that H-Y antibodies should inactivate Y-sperm but not X-sperm. Some investigators have claimed to skew mammalian sex ratios using methods based on this theory (McCormick et al., 1983; Boyce and Bennet, 1984). Bryant, in particular, has claimed a dramatic skewing of sex ratio using H-Y antibodies (Bryant, U.S. Pat. No. 4,191,749; Bryant, U.S. Pat. No. 4,448,767). However, experience has not borne-out these claims. Hoppe and Koo stated specifically that they were unable to skew sex ratio using antibodies against the H-Y antigen (Hoppe and Koo, 1984). As far as we know, no one has confirmed Bryant's claims.
There may be two reasons for the failure to confirm these results. First, the underlying theory appears to be wrong. The most recent evidence indicates that there is no difference in H-Y presence on mature X- or Y-sperm. While certain male tissues produce H-Y and express it as an integral membrane protein, Y-sperm do not appear to produce H-Y themselves. Rather, both X-and Y-sperm adsorb it to their surface (Garner, 1984). Hoppe and Koo have shown that both X- and Y-sperm react with H-Y antibody (Hoppe and Koo 1984). Our own evidence, which we present herein, corroborates this. Furthermore, Hoppe and Koo showed that as sperm mature, their ability to react with H-Y antigen declines, implying that H-Y is masked or lost from the sperm cell surface (Id.). Second, the experimental technique of some of these investigators may have been flawed. They based their conclusions on experiments with a limited number of animals, so that the sex ratios, while skewed, were not statistically significant (Moore and Gledhill, 1988). Therefore, there is no longer any reason to believe that one could successfully use the H-Y antigen to separate X-sperm and Y-sperm. Indeed, we are not aware of any methods currently in use which successfully use this strategy.
Fabricant et al., U.S. Pat. No. 4,722,887, refers to a method for separating X-and Y-sperm by polymeric phase separation based on differential expression of a sperm cell-surface sulfoglycolipid (SGG). However, the authors state that the evidence for sex-linked differences in this lipid is indirect--it is based on the sex-linked expression of enzymes which metabolize lipid substrates--and the authors expess only the expectation that SGG, itself, will prove to be sex-linked.
Another potential separation approach for X-sperm and Y-sperm is based on the known differences in the DNA contents of X-sperm and Y-sperm. Because the DNA content of X-sperm cells is greater than the DNA content of Y-sperm cells, investigators hoped that the respective live cell populations could be separated by density gradient sedimentation or flow cytometry. However, neither has proven to be possible.
One reason for this failure may be that the DNA content differences between X-sperm and Y-sperm are small. For example, the difference is believed to be only about 3.9% for bulls, 3.7% for boars and 4.1% for rams (Sumner 1971; Pearson, et al., 1973; Evans et al., 1972; and Gledhill, 1985). This translates into an approximate 0.003 difference in bouyancy--not enough to permit separation of whole sperm using available methods. While other mammals display somewhat higher differences in the relative DNA contents of X-sperm and Y-sperm, e.g., the vole (Microtus oregani) which has about a 9% difference, separation of whole sperm has also not been possible for these animals (see Pinkel et al., 1982). For example, investigators have tried to separate sperm based on their differing DNA content by density gradient sedimentation, but enrichment results could not be verified--one report claims to have slightly enriched a fraction of bull sperm, but not rabbit sperm (Schilling, 1971; Brandriff et al., 1986). Attempted separation on the basis of surface charge density imparted by DNA differences has also been inconsistent (Hafs and Boyd, 1971), or was based upon controversial quinacrine staining (Garner, 1984).
Another reason for these failures may be that the head, tail, and plasma membranes of the sperm, its other cellular material, and its highly compact nucleus all act to mask the small DNA content differences between X-sperm and Y-sperm. Some evidence for this masking effect is the fact that cytometric separation, while not feasible for whole sperm, has been useful to prepare enriched subpopulations of denuded sperm nuclei. Using this technique, the sperm nuclei are first separated from the membranes and other material of whole sperm. They are then stained and partially sorted using a flow cytometer (Johnson and Pinkel, 1986). The result has been nuclei subpopulations enriched for the X- and Y-chromosome.
Investigators have also used this cytometric technique to test the results of various attempts to separate the X- and Y-sperm of whole sperm (a non-enriched sperm population). The Lawrence Livermore National Laboratory and Oklahoma State University made a comparative study of several of the above-described "enrichment" approaches (Pinkel et al., 1985). They analyzed sperm separated by convection-counterstreaming-galvanizaton, albumin gradient, density gradient, electromotility, and anti-H-Y antibodies. The results: "In no case was enrichment of either sperm population observed." (Id. at p. 130.) This finding is consistent with other studies of attempted enrichment: albumin density gradient (Brandriff et al., 1986) and monoclonal anti-H-Y antibodies (Hoppe and Koo, 1984).
Monoclonal antibodies to sperm surface antigens which have heretofore been prepared also do not distinguish X- and Y-sperm. They bind to both X-sperm and Y-sperm, and either inactivate or immobilize both types of sperm cells (Schmell et al., 1982, and Peterson et al., 1981). Monoclonal antibodies appear to inhibit sperm-egg binding without regard to whether they bind to the sperm acrosome, head, midpiece, or tail. Further, antibodies specifically binding to the midpiece or tail have also been observed to immobilize sperm cells. (For a review of antibodies inhibiting fertility see Alexander and Anderson, 1987.)
Prior to this invention, it was not known whether one could isolate subpopulations of cells enriched in either X-sperm or Y-sperm. Nor was it known whether the plasma membranes of these enriched subpopulations would contain unique, sex selective constituents, such as proteins, glycoproteins, or lipoproteins.
In light of these failures, we decided to focus on the sperm cell surface as a possible tool for sperm separation. Studies of the cell membrane of unsexed mammalian sperm indicated that more than 1000 proteins are present on it. See, e.g., Noland et al. (1983 and 1984); (Russell et al., 1983); (Bradley et al., 1981); and (Hughes and August, 1981, and Crichton and Cohen, 1983). All of these studies used mixed membranes of both X- and Y-sperm. Therefore, they failed to distinguish between membranes and constituents characteristic of X-sperm-enriched subpopulations and those of Y-sperm-enriched subpopulations. Consequently, no one, until now, has been able to perform an analysis of X-sperm or Y-sperm membranes, to obtain usable quantities of whole cells enriched for X- or Y-sperm, to identify a sex-chromosome associated membrane protein of mammalian sperm, or to isolate such proteins.